|Publication number||US7180852 B1|
|Application number||US 11/240,105|
|Publication date||Feb 20, 2007|
|Filing date||Sep 30, 2005|
|Priority date||Dec 21, 2000|
|Also published as||US6982951, US7450496, US7688720, US20020097671, US20050232144|
|Publication number||11240105, 240105, US 7180852 B1, US 7180852B1, US-B1-7180852, US7180852 B1, US7180852B1|
|Inventors||Robert Duncan Doverspike, Charles Robert Kalmanek, Jr., Guangzhi Li|
|Original Assignee||At&T Corp.|
|Export Citation||BiBTeX, EndNote, RefMan|
|Patent Citations (8), Referenced by (31), Classifications (18), Legal Events (2)|
|External Links: USPTO, USPTO Assignment, Espacenet|
This application is a continuation of prior application Ser. No. 09/909,102 filed Jul. 19, 2001 now U.S. Pat. No. 6,982,951, which claims the benefit of U.S. Provisional Application No. 60/257,029, filed on Dec. 21, 2000, the contents of which are incorporated herein by reference.
This application is related to U.S. Utility patent application, “METHODS AND SYSTEMS FOR FAST RESTORATION IN A MESH NETWORK OF OPTICAL CROSS CONNECTS,” Ser. No. 09/474,031, filed on Dec. 28, 1999, which is incorporated by reference herein.
The invention relates to telecommunications networks, and more particularly to selecting restoration paths in a telecommunications network.
Modern telecommunication networks are reconfigurable and provide for fast restoration from network failures. For example, and in the context of optical networking, Synchronous Optical Network (SONET) rings provide the primary technology for optical layer communication and restoration from network failures. SONET rings tend to be capacity inefficient when compared to “mesh” topologies in networks with a high degree of connectivity and when, because of size limitations, connections are forced to route through many interconnected rings. As optical-cross connects (OXCS) are deployed within today's transport networks based on wavelength-division multiplexing (WDM), the potential emerges to provide on-demand establishment of high-bandwidth connections (also referred to in the art as “lightpaths”). Emerging standards such as Multi-Protocol Lambda Switching (“MPL(ambda)S”) provide a standardized optical network control plane that is essential for building an effective platform for vendor interoperability. See, e.g., D. Awduche et al., “Multi-Protocol Lambda Switching: Combining MPLS Traffic Engineering Control with Optical Crossconnects,” IETF Internet Draft, http://www.ietf.org/internet-drafts/draft-awduche-mpls-te-optical-01.txt (November 1999). Unfortunately, few recent contributions to the art have addressed the need for fast failure restoration in such networks.
In co-pending commonly-assigned U.S. Utility patent application, “METHODS AND SYSTEMS FOR FAST RESTORATION IN A MESH NETWORK OF OPTICAL CROSS CONNECTS,” Ser. No. 09/474,031, filed on Dec. 28, 1999, which is incorporated by reference herein, a restoration methodology is disclosed that utilizes pre-computed restoration routes disjoint from the normal communication path—but wherein the channels/wavelengths may be chosen dynamically during the restoration process. The invention therein disclosed can potentially provide restoration competitive with SONET ring restoration speeds. There is, however, still a need for a flexible and practical methodology for selecting an advantageous restoration path that may be utilized in a restoration process such as the one disclosed in the above application. Moreover, there is a need for a distributed approach to restoration that permits the information needed for restoration path selection to be distributed throughout the network with a minimum amount of signaling overhead.
A method of selecting a restoration path in a mesh telecommunication network is disclosed that advantageously is practical and flexible and may be pre-computed along with a service connection path during the setup of the connection. The restoration path is selected from a graph of links in the network which are physically diverse from the service path. For example, in the context of optical networking, the links do not share a common fiber span with the service path. Weights are computed for the links using an array representing a restoration link capacity—which is expressed as a number of channels/wavelengths in optical networking—needed on each link over possible failures of the service path. The restoration path is selected by minimizing the weights for each link in the restoration path. In accordance with another aspect of the invention, the information used to select the restoration path is advantageously distributed among nodes in the network. A source node selecting a service path through the network to a destination node sends a first message along the service path to the destination node, thereby setting up cross-connections for the service path. The destination node sends a second message back to the source node along nodes in the service path responsible for maintaining link information, whereby the nodes update the message with an array representing a restoration link capacity (channels/wavelengths) needed on each link over possible failures of the service path. The array is used by the source node to select the restoration path. The source node then sends a third message along the service path and the restoration path to reserve resources. This provides a distributed approach for establishment of restorable connections in a dynamically reconfigurable mesh network with a minimum amount of signaling overhead. Restoration paths are advantageously diverse from service connection paths while restoration link capacity advantageously may be shared among different connections that would re-route over the restoration path during non-simultaneous failures.
These and other advantages of the invention will be apparent to those of ordinary skill in the art by reference to the following detailed description and the accompanying drawings.
With reference to
Each OXC has an associated optical layer control function, integrated with the OXC or residing physically on a separate controller, which controls the cross-connections of the OXCs and which can be used to communicate with the other OXCS. See, e.g., co-pending commonly-assigned U.S. utility patent applications, “METHOD AND APPARATUS FOR ROUTING INFORMATION OVER OPTICAL AND ELECTRICAL PATHWAYS,” Ser. No. 09/685,953, filed Oct. 12, 2000; “CONTROL OF OPTICAL CONNECTIONS IN AN OPTICAL NETWORK,” Ser. No. 09/769,735, filed Jan. 26, 2001; and “CONTROL OF OPTICAL CONNECTIONS IN AN OPTICAL NETWORK,” Ser. No. 09/769,780, filed Jan. 26, 2001, which are incorporated by reference herein. It is assumed, without limitation, that the optical layer control function is distributed. The cross-connection controllers can communicate with one another across a dedicated signaling channel (or a separate signaling network) using known protocols such as Open Shortest Path Forwarding (OSPF). See, e.g. J. Moy, “OSPF Version 2,” IETF Network Working Group, RFC 2328 (April 1998); R. Coltun, “The OSPF Opaque LSA Option,” IETF Network Working Group, RFC 2370 (July 1998). OSPF link state advertisement (LSA) messages can be utilized to disseminate topology and channel usage information. It is advantageous to configure one OXC as a “master” OXC node for each fiber span and optical link, thereby allocating the control of the fiber span and optical link data among the OXCs in an unambiguous way. The master nodes are responsible for maintaining the link information. For each fiber span, if there is an OXC present at both terminating nodes of the fiber span, one can arbitrarily configure the OXC with lexicographically or numerically higher ID as the master. If there is only one OXC node present at a terminating node of the fiber span, then this node can be designated the master. If neither node has an OXC present, it is necessary to configure a neighbor OXC node as the proxy master of the fiber span. See tables in
It is assumed that mesh network 100 provides for restorable connections that may be established and torn down dynamically. Selecting optimal restoration parameters involves a variety of considerations and objectives.
First, it is advantageous to select a restoration path Pr that is physically “diverse” from the service connection path Ps. In other words, the restoration path and the service connection path should not belong to a group of links (referred to in the art as a “shared risk link group”) sharing some common infrastructure that could subject the links to a possible single failure, e.g. a backhoe cutting a single fiber conduit. Consider
Second, it is advantageous to share restoration channels on a common link of multiple restoration paths, in particular where fiber span failures are non-simultaneous. Unused channels may be reserved, i.e. not used for service paths, to ensure that adequate restoration capacity is available upon failure (alternatively, a dedicated restoration connection—referred to in the art as “1+1” protection—can be utilized although this tends to be impractical except for some high-priority services). A single restoration channel on a common link of multiple restoration paths can be shared by non-simultaneous fiber span failures. For example, a channel can be used on a link of restoration path Pr1 to reroute service path Ps1 due to failure of fiber span i along Ps1, That same channel can be used to reroute service path Ps2 to restoration path Pr2 due to failure of fiber span j along Ps2 as long as failures i and j do not occur simultaneously. On the path of Pr, enough channels should be reserved such that in any single fiber span failure, there are enough channels to restore all failed service paths. The total reserved channels, however, should be as small as possible.
In accordance with an aspect of the invention, the restoration path is selected in a manner that permits restoration channels to be dynamically assigned and shared when a given fiber span fails. The restoration path selection method can be utilized, without limitation, with a restoration architecture as disclosed in co-pending, commonly-assigned U.S. Utility patent application, “METHODS AND SYSTEMS FOR FAST RESTORATION IN A MESH NETWORK OF OPTICAL CROSS CONNECTS,” Ser. No. 09/474,031, filed on Dec. 28, 1999, which is incorporated by reference herein. As disclosed therein, pre-computed restoration paths may be stored at the endpoint nodes of the connection and utilized, upon a network failure, to reroute the service connection. Although described with particular reference to the restoration method described therein, one of ordinary skill in the art would readily recognize how to utilize the invention with other restoration methods. Requests to establish optical connections arrive at an OXC typically from a higher layer node (e.g., an Internet Protocol (IP) layer or element management system) via a User Network Interface (UNI). See, e.g. McAdams, L. and J. Yates, eds., “User to Network Interface (UNI) Service Definition and Lightpath Attributes,” OIF Architecture Group, OIF2000.61 (2000). Each optical connection request may include restoration options, such as whether the connection should be restorable if it fails due to a network failure and, if so, whether the channels of the restoration path are dedicated to this connection or shared among other connections that fail from other potential, non-simultaneous failures. If shared, the network resources can be used more efficiently because fewer total channels need to be reserved for restoration. For example, a connection request, V, can be represented in the form <source, destination, restoration-type, size> where source is the origin OXC ID, destination is the terminus OXC ID, restoration-type is the type of restoration capability required for the connection (e.g. 1+1, non-restorable, mesh-restorable), and size is the bandwidth needed for the connection (e.g. represented as the number of channels needed for the connection).
The process of computation of service path and restoration path for a connection request relies on the information about the availability of optical network resources and the path selection objective. A general heuristic is to create some cost metric and select a “minimum weight” path among all suitable paths that minimizes the cost metric and has the required size for the connection request. Additionally, several information metrics are involved in the path selection process when restoration resources may be shared between different restoration paths. For example, each link has a maximum number of installed channels, per the link augmentation (planning) processes: of that total, some channels are assigned to service paths while other channels are reserved for restoration paths. The remaining channels are unassigned and free to be allocated to new connections. Changes in the metrics need to be advertised, e.g. as part of LSAs in extended OSPF, so that accurate information is available to every node.
SERVICE PATH. Selecting a service path in response to the communication request, accordingly, may be accomplished by computing a path between the source and destination that minimizes some cost metric and which has the required size for the connection request. It is assumed that each OXC node has knowledge of the whole optical network topology and the number of free channels on each link as well as some optical link weight function. A known shortest path algorithm such as Dijkstra's shortest path algorithm may be used to compute the minimal weight path through the network.
RESTORATION PATH: Selecting a restoration path in response to a communication request requiring a mesh-restorable connection is more challenging. It should be noted that the restoration path may not be the “min weight” path in any typical sense, because it can share restoration resources with other restoration paths if shared restoration is required. Note that although to provide clarity with the example, E0 was defined above as the set of fiber spans, without loss of generality it can also represent any set of shared risk link groups (“SRLGs”) over which one wishes to provide restoration. It is advantageous to define the following parameters:
All of the information needed for the computation of the service and restoration paths could be maintained at every OXC node. This would require maintaining the entire two-dimensional array, failneedsk, at each OXC node. Whenever a new connection is provisioned, the matrix would need to be updated, e.g. by flooding of link state advertisement messages. In accordance with another aspect of the invention, however, portions of the array can be distributed around certain nodes whereby the array portions are updated along the restoration and service paths during connection establishment. This serves to avoid flooding and minimizes the storage requirements for this information. Thus, in accordance with a preferred embodiment of this aspect of the invention, each cross connect stores only the one-dimensional arrays failneeds• for each fiber span s for which it is a master node and failneed•k for each optical link kεE0 for which it is a master node.
As an example of the processing performed in
The foregoing Detailed Description is to be understood as being in every respect illustrative and exemplary, but not restrictive, and the scope of the invention disclosed herein is not to be determined from the Detailed Description, but rather from the claims as interpreted according to the full breadth permitted by the patent laws. It is to be understood that the embodiments shown and described herein are only illustrative of the principles of the present invention and that various modifications may be implemented by those skilled in the art without departing from the scope and spirit of the invention. For example, the detailed description has been presented particularly in the context of an optical networking architecture; however, the principles of the present invention could be extended to other cross-connect technologies. Such an extension could be readily implemented by one of ordinary skill in the art given the above disclosure.
|Cited Patent||Filing date||Publication date||Applicant||Title|
|US5218601 *||Dec 20, 1990||Jun 8, 1993||Fujitsu Limited||Method for searching for alternate path in communication network|
|US5812524 *||Oct 25, 1996||Sep 22, 1998||Mci Communications Corporation||Deterministic selection of an optimal restoration route in a telecommunications network|
|US5941992 *||Aug 13, 1997||Aug 24, 1999||Mci Communications Corporation||Distributed method and system for excluding components from a restoral route in a communications network|
|US5943314 *||Dec 30, 1996||Aug 24, 1999||Mci Communications Corporation||Method and system of distributed network restoration with multiple failures|
|US6282170 *||May 29, 1997||Aug 28, 2001||Nortel Networks Limited||Network restoration routing optimization|
|US6324162 *||Jun 3, 1998||Nov 27, 2001||At&T Corp.||Path-based restoration mesh networks|
|US6600719 *||Jan 4, 2000||Jul 29, 2003||At&T Corp.||Method and apparatus for restoring a network|
|US6725401 *||Oct 26, 2000||Apr 20, 2004||Nortel Networks Limited||Optimized fault notification in an overlay mesh network via network knowledge correlation|
|Citing Patent||Filing date||Publication date||Applicant||Title|
|US7451340||Sep 26, 2003||Nov 11, 2008||Lucent Technologies Inc.||Connection set-up extension for restoration path establishment in mesh networks|
|US7500013||Apr 2, 2004||Mar 3, 2009||Alcatel-Lucent Usa Inc.||Calculation of link-detour paths in mesh networks|
|US7529183 *||Aug 22, 2005||May 5, 2009||Alcatel||Method of setting up a backup path in a transport network|
|US7545736||Sep 26, 2003||Jun 9, 2009||Alcatel-Lucent Usa Inc.||Restoration path calculation in mesh networks|
|US7606237||Oct 20, 2009||Alcatel-Lucent Usa Inc.||Sharing restoration path bandwidth in mesh networks|
|US7643408||Mar 31, 2004||Jan 5, 2010||Alcatel-Lucent Usa Inc.||Restoration time in networks|
|US7646706||Jan 12, 2010||Alcatel-Lucent Usa Inc.||Restoration time in mesh networks|
|US7652983||Jan 26, 2010||At&T Intellectual Property Ii, L.P.||Method for restoration and normalization in a mesh network|
|US7689693||Sep 26, 2003||Mar 30, 2010||Alcatel-Lucent Usa Inc.||Primary/restoration path calculation in mesh networks based on multiple-cost criteria|
|US7941778||May 10, 2011||At&T Intellectual Property I, Lp||System and method of determining minimum cost path|
|US7974183||Dec 10, 2009||Jul 5, 2011||At&T Intellectual Property Ii, Lp||Method for restoration and normalization in a mesh network|
|US8111612||Feb 7, 2012||Alcatel Lucent||Link-based recovery with demand granularity in mesh networks|
|US8260896||Sep 4, 2012||Mwa Intelligence, Inc.||Monitoring business machines using a mesh network on field nodes|
|US8296407||Oct 23, 2012||Alcatel Lucent||Calculation, representation, and maintenance of sharing information in mesh networks|
|US8867333 *||Sep 26, 2003||Oct 21, 2014||Alcatel Lucent||Restoration path calculation considering shared-risk link groups in mesh networks|
|US9097555||May 8, 2012||Aug 4, 2015||Mwa Intelligence, Inc.||Monitoring device mesh network systems and methods|
|US20040190441 *||Sep 26, 2003||Sep 30, 2004||Alfakih Abdo Y.||Restoration time in mesh networks|
|US20040190445 *||Sep 26, 2003||Sep 30, 2004||Dziong Zbigniew M.||Restoration path calculation in mesh networks|
|US20040193724 *||Aug 12, 2003||Sep 30, 2004||Dziong Zbigniew M.||Sharing restoration path bandwidth in mesh networks|
|US20040193728 *||Sep 26, 2003||Sep 30, 2004||Doshi Bharat T.||Calculation, representation, and maintanence of sharing information in mesh networks|
|US20040205236 *||Mar 31, 2004||Oct 14, 2004||Atkinson Gary W.||Restoration time in mesh networks|
|US20040205237 *||Sep 26, 2003||Oct 14, 2004||Doshi Bharat T.||Restoration path calculation considering shared-risk link groups in mesh networks|
|US20040205239 *||Sep 26, 2003||Oct 14, 2004||Doshi Bharat T.||Primary/restoration path calculation in mesh networks based on multiple-cost criteria|
|US20040255765 *||Apr 2, 2004||Dec 23, 2004||Roland Corporation||Electronic percussion instrument|
|US20050147603 *||Dec 24, 2003||Jul 7, 2005||Smith Henry J.||Autoantibodies utilized as carrier agents for pharmaceutical compounds used in tumor imaging and cancer treatment|
|US20050220026 *||Apr 2, 2004||Oct 6, 2005||Dziong Zbigniew M||Calculation of link-detour paths in mesh networks|
|US20050226212 *||Apr 2, 2004||Oct 13, 2005||Dziong Zbigniew M||Loop avoidance for recovery paths in mesh networks|
|US20050240796 *||Apr 2, 2004||Oct 27, 2005||Dziong Zbigniew M||Link-based recovery with demand granularity in mesh networks|
|US20060045007 *||Aug 22, 2005||Mar 2, 2006||Alcatel||Method of setting up a backup path in a transport network|
|US20090132989 *||Nov 15, 2007||May 21, 2009||At&T Knowledge Ventures, Lp||Method of Determining Minimum Cost Path|
|US20100091647 *||Dec 10, 2009||Apr 15, 2010||Guangzhi Li||Method for Restoration and Normalization in a Mesh Network|
|U.S. Classification||370/217, 709/239, 379/221.04, 370/221, 370/225|
|International Classification||H04J14/02, H04L12/26, H04Q11/00|
|Cooperative Classification||H04J14/0241, H04J14/0227, H04J14/0295, H04Q2011/0088, H04Q11/0062, H04Q2011/0081, H04J14/0284|
|European Classification||H04J14/02N5, H04Q11/00P4, H04J14/02M|
|Jul 2, 2010||FPAY||Fee payment|
Year of fee payment: 4
|Jul 25, 2014||FPAY||Fee payment|
Year of fee payment: 8